Important Dates

Abstract: Is Artificial Intelligence intelligent? No, according to John Searle’s notorious Chinese Room argument. Consider Searle (who doesn’t understand Chinese) sitting in a room with a manual containing all the rules for answering questions in Chinese – the same rules that would be available to an AI. Whenever Searle receives a question in Chinese, he answers by blindly following the rules. Searle’s argument is that if he, who cannot understand Chinese, can produce outputs that are indistinguishable from a Chinese speaker, then AI has no understanding either. I assess Searle’s argument, analyse the major responses to him and consider recent developments in AI, such as deep learning. I propose a novel response that draws on research on the phylogeny and ontogeny of mind. Specifically, I argue that Searle's argument illicitly trades on a homuncular and anthropocentric approach to intelligence. Instead, I offer ways to analyse intelligence in systems where there is no single locus of understanding. My approach allows us to address and unify the study of intelligence in a range of entities as diverse as AI, cephalopods and forms of ‘minimal cognition’ (e.g. slime moulds) – all of which can be productively analysed as forms of distributed intelligence. In doing so, I use AI as one case to challenge established notions of intelligence, while engaging with AI as part of a larger research program of how to think about intelligence. [Poster]